From 0fae7d96fb219dde7615efa82da6eb281f043340 Mon Sep 17 00:00:00 2001
From: Logan Williams <logan@bellingcat.com>
Date: Tue, 9 May 2023 12:12:02 +0200
Subject: [PATCH] Detect running in docker container in WACZ enricher

---
 Dockerfile                                   |  4 +-
 src/auto_archiver/enrichers/wacz_enricher.py | 53 +++++++++++++-------
 2 files changed, 39 insertions(+), 18 deletions(-)

diff --git a/Dockerfile b/Dockerfile
index 60805cd..fef93b5 100644
--- a/Dockerfile
+++ b/Dockerfile
@@ -1,6 +1,8 @@
 # stage 1 - all dependencies
 FROM webrecorder/browsertrix-crawler:latest
 
+ENV RUNNING_IN_DOCKER=1
+
 WORKDIR /app
 
 # TODO: use custom ffmpeg builds instead of apt-get install
@@ -28,7 +30,7 @@ COPY ./src/ .
 # TODO: figure out how to make volumes not be root, does it depend on host or dockerfile?
 # RUN useradd --system --groups sudo --shell /bin/bash archiver && chown -R archiver:sudo .
 # USER archiver
-ENTRYPOINT ["python"]
+ENTRYPOINT ["python3"]
 # ENTRYPOINT ["docker-entrypoint.sh"]
 
 # should be executed with 2 volumes (3 if local_storage)
diff --git a/src/auto_archiver/enrichers/wacz_enricher.py b/src/auto_archiver/enrichers/wacz_enricher.py
index f594efe..c4530ed 100644
--- a/src/auto_archiver/enrichers/wacz_enricher.py
+++ b/src/auto_archiver/enrichers/wacz_enricher.py
@@ -26,30 +26,49 @@ class WaczEnricher(Enricher):
 
     def enrich(self, to_enrich: Metadata) -> bool:
         # TODO: figure out support for browsertrix in docker
+
         url = to_enrich.get_url()
 
         if UrlUtil.is_auth_wall(url):
             logger.debug(f"[SKIP] SCREENSHOT since url is behind AUTH WALL: {url=}")
             return
-
-        logger.debug(f"generating WACZ for {url=}")
+        
         collection = str(uuid.uuid4())[0:8]
         browsertrix_home = os.path.abspath(ArchivingContext.get_tmp_dir())
-        cmd = [
-            "docker", "run",
-            "--rm",  # delete container once it has completed running
-            "-v", f"{browsertrix_home}:/crawls/",
-            # "-it", # this leads to "the input device is not a TTY"
-            "webrecorder/browsertrix-crawler", "crawl",
-            "--url", url,
-            "--scopeType", "page",
-            "--generateWACZ",
-            "--text",
-            "--collection", collection,
-            "--behaviors", "autoscroll,autoplay,autofetch,siteSpecific",
-            "--behaviorTimeout", str(self.timeout),
-            "--timeout", str(self.timeout)
-        ]
+        
+        if os.getenv('RUNNING_IN_DOCKER'):
+            logger.debug(f"generating WACZ without Docker for {url=}")
+
+            cmd = [
+                "crawl",
+                "--url", url,
+                "--scopeType", "page",
+                "--generateWACZ",
+                "--text",
+                "--collection", collection,
+                "--behaviors", "autoscroll,autoplay,autofetch,siteSpecific",
+                "--behaviorTimeout", str(self.timeout),
+                "--timeout", str(self.timeout)
+            ]
+        else:
+            logger.debug(f"generating WACZ in Docker for {url=}")
+            
+            cmd = [
+                "docker", "run",
+                "--rm",  # delete container once it has completed running
+                "-v", f"{browsertrix_home}:/crawls/",
+                # "-it", # this leads to "the input device is not a TTY"
+                "webrecorder/browsertrix-crawler", "crawl",
+                "--url", url,
+                "--scopeType", "page",
+                "--generateWACZ",
+                "--text",
+                "--collection", collection,
+                "--behaviors", "autoscroll,autoplay,autofetch,siteSpecific",
+                "--behaviorTimeout", str(self.timeout),
+                "--timeout", str(self.timeout)
+            ]
+
         if self.profile:
             profile_fn = os.path.join(browsertrix_home, "profile.tar.gz")
             shutil.copyfile(self.profile, profile_fn)